Rolling out M365 Copilot securely: permissions first

The fear around Microsoft 365 Copilot is almost always aimed at the wrong target. What people fear is the AI: that it hallucinates, that it carries company data outside. What then happens in the pilot is far more mundane. Someone asks Copilot about salaries and gets a summary from a SharePoint folder that has been shared with "Everyone" for years. If you want a secure M365 Copilot rollout, you rarely have an AI problem. You have a permissions problem, and it is older than Copilot. Data protection and security with Copilot hinge on two things that have nothing to do with the language model: who is allowed to open which repository today, and where that data sits.

What Copilot does and does not do

Microsoft states it plainly in its own documentation: "Microsoft Copilot only surfaces organizational data to which individual users have at least view permissions." (Source: Microsoft Learn, Data, Privacy, and Security for Microsoft Copilot)

In essence: Copilot does not invent access. It reads exactly what the signed-in user could already open today, across Mail, Teams, SharePoint and OneDrive. The same documentation also states that prompts and responses are not used to train the language models, and that encryption via sensitivity labels is respected. The same principle applies to extensions: if you connect Copilot to other systems via Graph connectors or agents, it only returns what the user has access to there as well. The boundary runs along permissions, nowhere else. What happens when someone tries to exploit that boundary through a planted document is covered in prompt injection explained.

For the question of whether the tool itself is built cleanly, that is good news. And this is exactly where the thinking error sits that we see most often in conversations about AI adoption: "Copilot respects your permissions" is only as reassuring as your permissions themselves.

Is M365 Copilot compliant with data protection law?

Short answer: the product brings the prerequisites, the compliance happens at your end. Tenant isolation, encryption and the quote above are documented. Under the Swiss FADP your company stays responsible for processing the personal data all the same, and two things on your side decide more than any assurance from Microsoft: who may open which repository, and where your data is stored.

Since September 2026, the storage location is a switch someone has to set deliberately. Microsoft introduced a setting in the admin center called "Store Microsoft 365 customer data in-country", found under Settings, Org settings, Organization profile, Data location, Commitments. It covers Exchange Online, SharePoint, OneDrive, Teams as well as M365 Copilot and Copilot Chat. Set to On, your data at rest stays in your tenant's country. Set to Off, Microsoft may store it regionally within the EU Data Boundary. (Source: Microsoft Learn, Product Terms Data Residency)

Two dates belong to this. The switch has been visible since 14 September 2026, but nothing moves yet. On 14 December 2026 the recorded choice takes effect, and for tenants left on Off, Microsoft may then begin moving data regionally. The documentation is clear on both points: for tenants created after 14 September 2026 the setting is Off by default, and for older tenants the configuration sits in the Message Center. The choice is not permanent. You can switch it back, and Microsoft says moving the data back takes around six months.

For most SMEs this is no drama, because the data stays in Europe. It is, however, the question a customer asks in a security questionnaire, and the one a contract naming Switzerland as the storage location has to answer. Not answering it actively is also an answer. The look at the admin center therefore belongs in the same week as the licence order. One aside: not every tenant sees the switch. It appears only for a default geography of Switzerland, France, Germany, Norway or Sweden, and not for tenants that bought Advanced Data Residency or Multi-Geo.

Locked is not the same as unfindable

Before Copilot, what protected you in practice was not your permissions concept. What protected you was obscurity.

Hardly anyone voluntarily clicks through tens of thousands of documents in other teams' sites. The salary overview in the wrong folder, the due diligence memo from an old project, the termination list from the reorganization: all theoretically accessible, practically invisible. An M365 tenant that has grown for ten years carries migration archives, orphaned Teams channels and shares set to "Everyone except external users" that were forgotten long ago.

Copilot changes that. It searches semantically across everything the user has access to and delivers the most relevant results in seconds, neatly summarized. "Theoretically accessible" becomes "one question away". The tool leaks nothing. It makes visible what was already open.

In our experience, this is the moment when a Copilot rollout stalls: not because of the technology, but because the first serious test surfaces things that should never have been lying there. That is uncomfortable. It is also the most honest inventory of your tenant you have ever received for free.

The pattern usually looks the same. A company starts the pilot with a handful of people from IT and the business units. In the first days no catastrophes appear, but small things with weight do: an old management presentation with salary bands, filed in a project site from five years ago. A quotations folder that half the company can access because the project team invited "everyone" back then and the site was never closed. None of this is a Copilot defect. All of it had been open for years. What is new is that someone now sees it, and that management has a concrete list for the first time instead of a vague gut feeling.

Why another tool is the wrong first answer

The reflex at this point is familiar: buy another product. The market now offers an add-on for almost every Copilot worry, from oversharing scanners to AI monitoring suites.

Some of them are useful. But a tool that reports oversharing does not clean up your shares. It only documents that they are open, and sends you an annual invoice for it. The actual work stays the same and is unglamorous: know which data would hurt you, clean up the repositories holding exactly that data first, name owners, use sharing links with expiry dates instead of forever. That costs working time and discipline, not licenses. How to identify data worth protecting in the first place is covered in our post on data classification; it is the foundation here too.

On top of that: you already pay for part of the tooling for this cleanup through the Copilot licence itself. Microsoft puts it plainly: SharePoint Advanced Management "is included with your Microsoft Copilot license". (Source: Microsoft Learn, Secure and Governed Data Foundation for Microsoft Copilot) It contains exactly the reports this work needs: a snapshot of how broadly permissions are spread across all sites, the sharing links created in the last 28 days, a list of the content shared with "Everyone except external users", and site access reviews you can delegate to the owners of the affected sites. Plus Restricted Content Discovery: a per-site switch that keeps sensitive content out of Copilot and organization-wide search without changing access for the site's members. That is the relief valve for the one site you will not clean up before the rollout, not a substitute for cleaning up. Unspectacular, and no sales rep calls you about it, because it is already paid for. That is why we start there. Exhaust what you have before talking about anything new: the same logic we apply to security tools in general, and it applies to AI projects unchanged.

Then there is the Swiss context, because it often causes confusion: Switzerland has no AI act of its own. On 12 February 2025 the Federal Council decided to ratify the Council of Europe's AI Convention, and according to BAKOM the consultation is planned to open in spring 2027. (Source: BAKOM, Federal Council takes note of the interim status of AI regulation) For your rollout that means there is no date you can wait for. We wrote up the full timeline in Switzerland and its AI law. What applies today is the FADP. If Copilot pulls personal data together in its answers, that is a data protection matter under existing law, not a new legal field. And here too: the duty to make personal data accessible only to those who need it existed before Copilot. Copilot only makes it measurable whether you fulfil it. And the responsibility stays with you as a company, not with Microsoft, however well the platform is documented.

Banning it is the most expensive option

The opposite reflex to buying add-ons is blocking: Copilot stays off until the tenant is perfectly tidy. That feels cautious and creates the bigger problem in practice. The demand for AI does not disappear; it migrates into private ChatGPT accounts in the browser. There, company data actually leaves the tenant, with no permission model, no log, no sensitivity label. We have known this effect for years from shadow IT; with AI it just repeats faster and with more sensitive data. A blocked Copilot next to freely reachable browser AIs is the worst combination of both: no productivity gain, and the data risk sits exactly where you see it least. If you would rather meet that demand with a company plan, we compared ChatGPT Team and ChatGPT Enterprise on price and administrative effort.

The tenant does not have to be perfect before the start either. It has to be clean in the places where it hurts, and the rest improves during operation. Perfection as a launch condition is just a polite form of a ban.

A secure Copilot rollout: what must be in place first

We advise nobody to put Copilot on ice because of this. The productivity gain is real, and your vendor's competitors are no better, they are just less deep in your data. In our view, the only open question is the order of steps. Four things belong before the broad rollout:

  1. You know which data is critical. Not as a 40-page classification concept, but as a short, honest list: salaries, health data, M&A, customer contracts, price calculations. What is on that list determines where you look first.
  2. The critical repositories are checked. Not the whole tenant, nobody manages that before a rollout. But the sites and teams holding the list from point one have clarified access and an owner who has to maintain it.
  3. A pilot that actively searches. Give a small group Copilot with the explicit task of finding things they should not see. The prompts may sound banal: What does management earn? Summarize the recent terminations. What terms does our biggest customer get? Every hit is a permission you had to fix anyway, except now you know about it.
  4. A short usage rule. Who gets Copilot, for what, and what still does not belong in a prompt. One page is enough to start. How such a rule fits into existing governance is covered in our post on AI governance. Retention belongs in the same rule: what M365 Copilot stores and how long it stays is your decision, not the product's.

After that, roll out in waves instead of everyone at once. Each wave brings new users with different access profiles and thereby uncovers different corners of the tenant; the rollout doubles as an ongoing permissions test you would otherwise pay dearly for. Exactly this order, data before licenses, is also the core of our work on AI governance and secure AI adoption.

The question before the first prompt

If you are the person who has to explain this to management, a renaming helps: do not request a budget for "AI security", which sounds like another bottomless pit. Request the permissions cleanup your tenant needed anyway, with Copilot as the occasion and the deadline. The difference is more than rhetoric: the work pays into every audit, every insurance questionnaire and every future incident, not just into a single tool. The question "What does the AI cost us on top?" becomes "What do we finally clean up while we are at it?".

Introducing Copilot is, in the end, a rare opportunity: for the first time there is a business reason to tackle the permissions chaos that has been known internally for years and never had priority. The companies that get the rollout right are AI-ready afterwards, and beyond Copilot: the same clarified data situation carries every further AI consulting engagement and every tool that comes after. And their tenant is, as a side effect, a smaller target for every other attack path, without a single new security product.

If you want an honest outside view of your data situation and permissions before the rollout: an intro call costs nothing and does not end in a shopping list.

And until then, the one question that sums up this whole article: Which file in your tenant would you never want to see in a Copilot answer, and do you know today who has access to it?

Frequently asked questions

Is M365 Copilot compliant with data protection law?

The product brings the prerequisites: tenant isolation, encryption and no use of prompts for model training. Under the Swiss FADP, your company stays responsible for processing the personal data. Two things on your side are therefore decisive: who may open which repository today, and where your data is stored.

Is our data used to train Copilot?

No. Microsoft states in its own documentation that prompts and responses from Microsoft 365 Copilot are not used to train the language models, and that encryption via sensitivity labels is respected. That does not hold for private AI accounts in the browser, where the provider's terms decide.

Does Copilot data stay in Switzerland?

Only if someone sets the switch. Microsoft introduced a setting in the admin center called "Store Microsoft 365 customer data in-country", and it covers Copilot and Copilot Chat too. Left on Off, Microsoft may store your data at rest regionally within the EU Data Boundary from 14 December 2026.

How do we stop Copilot from surfacing confidential files?

Through permissions first, because Copilot only shows what the signed-in user could open anyway. For the repositories that will not be cleaned up before the rollout there is Restricted Content Discovery: a per-site switch that keeps that site's content out of Copilot and organization-wide search without changing access for its members.